[Paper Review] On the Liveliness of Artificial Life

Yong Zher Koh, Maurice HT Ling|arXiv (Cornell University)|Feb 19, 2023

Origins and Evolution of Life54 references3 citations

TL;DR

This paper proposes a comprehensive definition of life as a system with a protective boundary, a self-sustaining program capable of improvisation, adaptation, and regeneration through non-interfering reactions. Applying this definition, the authors argue that certain digital organisms—though not all—can be considered alive, and introduce a continuity-based experimental framework to test future definitions of life.

ABSTRACT

There has been on-going philosophical debate on whether artificial life models, also known as digital organisms, are truly alive. The main difficulty appears to be finding an encompassing and definite definition of life. By examining similarities and differences in recent definitions of life, we define life as "any system with a boundary to confine the system within a definite volume and protect the system from external effects, consisting of a program that is capable of improvisation, able to react and adapt to the environment, able to regenerate parts of it-self or its entirety, with energy system comprises of non-interference sets of secluded reactions for self-sustenance, is considered alive or a living system. Any incomplete system containing a program and can be re-assembled into a living system; thereby, converting the reassembled system for the purpose of the incomplete system, are also considered alive." Using this definition, we argue that digital organisms may not be the boundary case of life even though some digital organisms are not considered alive; thereby, taking the view that some form of digital organisms can be considered alive. In addition, we present an experimental framework based on continuity of the overall system and potential discontinuity of elements within the system for testing future definitions of life.

Motivation & Objective

• To resolve the philosophical ambiguity over whether artificial life models (digital organisms) can be considered truly alive.
• To address the lack of a universally accepted, encompassing definition of life that applies to both biological and artificial systems.
• To propose a formal, operational definition of life that includes boundary, programmatic improvisation, adaptation, regeneration, and self-sustaining energy systems.
• To establish criteria under which incomplete systems can still be considered alive if they can be reassembled into living systems.
• To develop a testable experimental framework based on system continuity and component discontinuity for evaluating future definitions of life.

Proposed method

• Defining life as a system with a physical boundary confining its volume and shielding it from external interference.
• Specifying that a living system must contain a program capable of improvisation, environmental reaction, and adaptation.
• Requiring the system to regenerate parts or its entirety, indicating self-repair and persistence.
• Incorporating a non-interference set of secluded reactions as the energy system to ensure self-sustenance.
• Extending the definition to include incomplete systems that can be reassembled into a living system, thus preserving their potential for liveliness.
• Proposing an experimental framework based on continuity of the overall system and potential discontinuity of internal components to test the validity of the definition and future definitions of life.

Experimental results

Research questions

• RQ1 What criteria must a system meet to be considered alive, especially when it is artificial or digital?
• RQ2 Can digital organisms—despite being non-biological—qualify as living under a robust, formalized definition of life?
• RQ3 How can we distinguish between systems that are merely complex and those that are truly alive, especially when life is not universally defined?
• RQ4 In what conditions can an incomplete system still be considered alive if it can be reassembled into a functional, self-sustaining system?
• RQ5 What experimental framework can be used to evaluate and validate future definitions of life in artificial systems?

Key findings

• The proposed definition of life includes a boundary, a self-sustaining program capable of improvisation and adaptation, and a non-interfering set of reactions for energy production.
• Digital organisms that meet the criteria—particularly those capable of regeneration and environmental adaptation—can be considered alive under the proposed definition.
• Not all digital organisms are alive, but some qualify as living due to their structural and functional alignment with the definition.
• The framework for testing definitions of life is based on the continuity of the overall system and the potential discontinuity of its components, enabling empirical evaluation.
• The definition accounts for incomplete systems that can be reassembled into living systems, thereby preserving their status as potentially alive.
• The study provides a foundation for evaluating artificial life systems using a formal, testable, and philosophically grounded criterion for liveliness.